Endocrine therapy began concurrently with RT in both chemotherapy-receiving and chemo-naïve patients and mostly started slightly before the start of RT. The start date of endocrine therapy was not documented, but a later phone query to tamoxifen-users confirmed that the therapy was started, by most, before the start of RT (I).
5.5.1 Endocrine therapy, biomarkers and echocardiography in the patients who received chemotherapy and radiotherapy
The effect of endocrine therapy on the HA, ADMA and hscTnT levels, and echocardiography was explored in publication I. During chemotherapy, the HA levels decreased in a similar fashion in the 6 tamoxifen users and the 13 AI users, since endocrine therapy was started after the completion of chemotherapy.
However, the HA levels returned to baseline levels only in tamoxifen users (p
=0.753) and not in AI users (p=0.001). Additionally, the hscTnT release occurred
was prevented in the tamoxifen users during RT (p=0.750). A significant release continued in the AI users (p=0.006). The ADMA levels were not affected by endocrine therapy.
5.5.2 Endocrine therapy, biomarkers and echocardiography in chemo-naïve patients
In chemo-naïve patients, the associations among endocrine therapy, echocardiography and the biomarkers HA, ADMA, hscTnT, TGF-β1, PDGF and ST2 were explored in publications I, III and IV.
The baseline levels of HA were similar in AI users and in those without endocrine therapy and were therefore compared as one group to tamoxifen users.
Tamoxifen users had a higher baseline HA levels than nonusers (p=0.017), and they exhibited a significant increase in the HA level during RT (p=0.028). Non-users had stable HA levels (p=0.231). A trajectory analysis was also performed, and three groups were identified. Patients within group 1 had the highest HA levels, and patients within group 3 had the lowest HA levels. The OR for tamoxifen use was 11.71 (95% CI 1.79-76.55) between groups 1 and 3. There was a tendency for the ADMA levels to be lower in tamoxifen users than in nonusers (p=0.050) before RT, and this difference became significant by the end of RT (p=0.044). The hscTnT levels were unaffected by endocrine therapy during RT in the chemo-naïve patient population.
There were no significant differences in the TGF-β1 and ST2 levels before RT, after RT or at the three-year follow-up, or in the changes between these time points, when patients were grouped according to tamoxifen use or AI use (III, IV).
The PDGF levels, on the other hand, were significantly lower in tamoxifen users than nonusers (p=0.041) (III).
The echocardiographic parameters according to tamoxifen use, AI use and no endocrine therapy were explored in publication I. Tamoxifen users had fewer changes in echocardiographic parameters than AI users or patients without endocrine therapy. In tamoxifen users, only a significant decrease in CVIBS (p=0.046), a marker of LV function, was observed. AI users had a significant decline in LV function, measured by GLS (p=0.001), RV function, measured by TAPSE (p<0.001) and diastolic function, measured by mitral E (p=0.003). The patients without endocrine therapy had changes in CVIBS (p=0.001) and TAPSE
(p=0.028) as well as a thickening of the IVS (p=0.023) and PW (p=0.046) and an increased echodensity in scIBS (p=0.002) (publication I, Table 3).
In multivariable linear regression analyses, AI use was an independent predictor of GLS decline over the three-year follow-up. In an analysis with the TGF-β1 trajectory group, left-sided breast cancer, AI use and age, the GLS decline was independently predicted by all factors except age (III). In another analysis that included the change in the ST2 level over the three-year follow-up, AI use, laterality of breast cancer, age, BMI and hypertension as variables, AI use was the only significant predictor of the GLS decline over the three-year follow-up (IV).
6 DISCUSSION
We observed changes in several different cardiac biomarkers during chemotherapy, during RT and during the three-year follow-up in conjunction with the cardiac structural and functional changes evaluated by echocardiography. Endocrine treatments also affected both the biomarker levels and echocardiographic findings.
6.1 Changes in biomarkers and echocardiography in patients who were treated with chemotherapy and RT
We found a chemotherapy-induced decrease in HA levels in the 19 patients who were treated with chemotherapy. Furthermore, the HA level increased after chemotherapy during the completion of the RT course. However, the HA level did not return to its baseline level. A significant increase in hscTnT was also observed (I). Simultaneously, in the 11 left-sided breast cancer patients who also had echocardiography performed, a thickening of the LV, an increasing LVESD and PW were observed. Additionally dysfunction of the myocardium, namely a decrease in CVIBS, was observed. Due to the small sample size, we could not determine the associations between HA levels and echocardiographic changes in this part of the study (I).
To date, our study is the only one in which the effect of cancer treatments on HA has been evaluated. A possible explanation for the decrease in HA levels during chemotherapy is through the NO pathway. Most of our patients received anthracycline-based chemotherapy, the cardiotoxicity of which is at least partially mediated through anthracycline’s ability to activate iNOS and induce NO production, which then leads to the production of peroxynitrite, a cardiomyocyte damaging compound (117). HA, on the other hand, is thought to act as a substrate for NOS (246) and therefore, the decrease in HA levels could represent the increased use of HA for NOS activation.
The TnT release caused by chemotherapy has also been reported in earlier studies (217,230,231). The increase in TnI levels is even more widely reported
(174,226–229). The measurement of troponins is suggested by several guidelines as an aid to identify patients at risk for cardiotoxicity during and after treatments (7–
11).
The chemotherapy-induced changes in LV measurements are not a widely studied, but in an earlier study, increased LVESD and LVEDD at baseline seemed to be predictive of the development of major cardiac adverse events in patients receiving anthracyclines for various cancers (191). Furthermore, anthracyclines in children and in lymphoma patients and high-dose 5-FU for various cancers have been reported to induce a decrease in CVIBS, but no studies on breast cancer patients were found in the literature search (197–199). However, the significance of these findings that occurred early after the treatments in a small patient population requires further study.